Touch interfaces can be found in laptop computers, gaming devices, automobile dashboards, kiosks, operating rooms, factories, automatic tellers, and a host of portable devices such as cameras and phones. Touch interfaces provide flexible interaction possibilities that discrete mechanical controls do not. But prior art touch interfaces sacrifice an important part of the human experience: haptics. “Haptics” refers to the perceptual system associated with touch. Haptics lets us touch type, find a light switch in the dark, wield a knife and fork, enjoy petting a dog or holding our spouse's hand. Haptics is not just about moving one's hands, but it is about feeling things, recognizing objects (even without looking at them), and controlling the way that we interact with the world.
Haptics in the form of vibration is a familiar feature of electronic products such as pagers, cell phones, and smart phones. Although vibration has long been used as a silent ringer or alarm, it is increasingly used to provide tactile feedback to the human hand (especially the fingertips) when using a touch surface, such as a touch screen. Immersion Corporation, for instance, provides a number of hardware and software solutions for vibration-based haptic feedback. Their technology is considerably more advanced than what was traditionally used in pagers. They make use of piezoelectric actuators to enable high bandwidth control of vibration profiles. Nonetheless, their approach has certain drawbacks. For instance, the entire device vibrates so that any effect is felt in the hand holding the device as well as at the fingertip touching the touch surface or screen. Furthermore, it does not support multi-point haptics: because the entire device vibrates, each fingertip touching the screen experiences the same effect.
Recently, electrostatic actuation has been explored as a means to generate vibrations localized to the fingertip. Senseg Ltd. makes use of electrostatic forces to create vibrations of the fingertip that enable one to detect a variety of textures on a touch surface. Senseg's technology has the advantage that it generates no mechanical vibrations except at the surface of the skin. While their technology also has the potential to support multi-point haptics simply by using multiple electrodes on the same surface of a screen, in practice this is difficult to do. One reason is that it is difficult to make low-resistance electrical connection to electrodes that are not near the edge of a transparent screen. Thus, of the multiple electrodes, the ones not near the edge are slow to charge. Another reason is that the haptics must co-exist with some means of sensing fingertip locations. The most common technique for multi-touch sensing is “projected capacitive” sensing, which also makes use of electrostatic charges. To minimize the interaction between the electrostatic haptics and the projected capacitance sensing, the Senseg solution makes use of a single electrode for haptics, the size of the whole touch screen.
Multi-Point Haptics
A co-pending patent application by the present inventors (U.S. patent application Ser. No. 13/468,818, entitled Electrostatic Multi-touch Haptic Display) describes a number of ways of achieving multi-point electrostatic haptics. Certain aspects of that disclosure are noted herein as a background. For instance, the basis of electrostatic haptics is the modulation of frictional force via an electric field. The electric field is established at the point of contact between the fingertip and the touch surface. This is accomplished by placing one or more electrodes (haptic devices) on the touch surface of the substrate and insulating those electrodes from the fingertip with a dielectric layer. To set up an electric field, a circuit must be closed through the fingertip. There are two principal ways of doing this.
In the prior art, others have taught the method shown in FIG. 1a, which is a figure from U.S. Pat. No. 7,924,144, wherein capacitance of a finger-dielectric-electrode system is part of a circuit that is closed through a second contact at some other part of the body. Thus, FIG. 1a shows an apparatus which implements a capacitive electrosensory interface, having an electrical circuit that is closed between two separate contact locations, wherein both of the two locations are fingertips.
The present inventors have devised an alternative method shown in FIG. 1b, which is similar to a figure from U.S. patent application Ser. No. 13/468,695, entitled Touch Interface Device And Method For Applying Controllable Shear Forces To A Human Appendage, wherein two separate electrodes E and E′ (haptic devices) are covered by an insulating layer L and would be placed on a front or top surface of a substrate (not shown) at a single contact or touch location. The circuit is therefore closed through a single touch of a fingertip itself, not involving the rest of the body. This has the benefit of not requiring involvement of some other part of the body, but it has another benefit as well, which will be discussed herein.
To apply the two-electrode technique, it is necessary to create a suitable array of electrode pairs on the touch surface. As illustrated in FIG. 2, one approach to accomplish this arrangement for an apparatus, such as a mobile device 2, would be to simply tile a top surface or touch surface 4 with electrode pairs 6 that include electrodes 8 and 10. This top layer of electrodes has the advantages that electrodes 8, 10 can be placed precisely where they are needed on the surface 4 and that all electrodes can potentially be patterned from the same conductive layer. It will be appreciated that wires can be patterned from the same conductive material as the electrodes, or can be made of higher conductivity material.
However, this configuration has the disadvantage present in some prior art with respect to the need for respective thin conductive traces 14, 16 to connect to many of the electrodes, such as those that are not near an edge. Thin conductive traces 14, 16 with sufficiently low resistivity can be difficult to produce, especially if they need to be transparent to meet other design objectives. Another potential difficulty with this approach is that the electrode count may become quite large, especially as the touch surface becomes larger. If the x-axis requires N electrodes and the y-axis requires M, then the total electrode count with pairs, as shown in FIG. 2, is 2*M*N. Nonetheless, patterns like this one that tile the surface with electrode pairs may be used in certain situations, such as with devices having smaller screen sizes.
A second approach to creating an electrode array for the touch surface of an apparatus is shown in FIG. 3a and is referred to as a “Lattice.” The diagram in FIG. 3a focuses on the electrode array, for ease of understanding. While a pattern in the form of a lattice network of lines of diamond-shaped electrodes is shown, such a pattern and shape of electrodes need not be used, but the emphasis is on covering the surface (here shown as being generally planar) with N*M electrodes that can serve in pairs. In this figure, electrodes 20 run along or parallel to a first axis (for example the x-axis), and electrodes 22 run along or parallel to a second axis (for example the y-axis). The region where a given y-axis electrode 22 crosses a given x-axis electrode 20 defines a two-electrode region (like that shown in FIG. 1b) where electrostatic forces can be applied to a user's skin, such as to a fingertip.
As shown in FIG. 3a, any electrode 20 (x-axis) and electrode 22 (y-axis) can form a pair. If different voltages are applied to, for example, the electrodes 20 and 22, then an intersection of the respective electrodes 20, 22 becomes an active region or location where a finger will experience increased electrostatic force. In practice, AC voltages may be used and maximum electrostatic forces are produced when the two voltages are 180 degrees out of phase with one another.
The magnitude of the electrostatic force can be modulated in various ways. As a few examples, one approach is to change the magnitude of the voltages applied to the electrodes. Another is to vary the duty cycle of the voltage waveforms applied to the electrodes. Yet another is to control applied voltage or current based on a measure of the electrical charge on the electrodes 20, 22. A further approach is to vary the phase relationship between the voltages on the two electrodes 20, 22. The electrostatic force is maximized when the voltages on the two electrodes 20, 22 are completely out-of-phase with one another, and minimized when they are in phase because the circuit then is no longer closed locally through the touch, contact or engagement of a user's finger, such as at a fingertip, but must be closed through the capacitance of the rest of the user's body. If Cf is the capacitance from the finger to the electrodes and Cb is the capacitance from the rest of the body to the device ground, then the attenuation factor (ratio of force when touching in-phase electrodes to force when touching out-of-phase electrodes) is:
  Attenuation  =            (                        C          b                                      C            b                    +                      C            f                              )        2  
Normally, Cf is significantly larger than Cb (at least by a factor of 5), so the attenuation factor is quite significant: more than an order of magnitude.
A pattern in the form of a Lattice network or configuration also supports multi-point haptics to a certain extent. This is illustrated in FIG. 3b, which shows a pattern having a lattice network of electrodes that further includes electrodes 24 that run along or parallel to a first axis (for example the x-axis), and electrodes 26 that run along or parallel to a second axis (for example the y-axis), and in which the intersections between electrodes 20 and 22, as well as trace intersections between electrodes 24 and 26, each are used to define or control the electrostatic force acting on two respective fingers, with a first fingertip F represented by a first oval and a second fingertip FF represented by a second oval. There are, however, finger locations where this multi-point capability may break down. If, for instance, two fingertips lie on the same electrode, then it is difficult to apply very different forces to the two fingers. The reason for this is that the finger-to-finger impedance through the user's body is quite small relative to the electrode-to-finger impedance (1/(f Cf)) where f is the frequency of AC excitation. Thus, for instance, a second finger on an active x-axis electrode still has the benefit of the active y-axis electrode under the first fingertip. Note that in the example, x and y could be reversed. To ensure that the force on each finger is independent of the force on each other finger, it is necessary that they be parts of different circuits. As described above, this could be accomplished by the arrangement in FIG. 2, but with drawbacks.
Multi-Touch Sensing
Most modern multi-touch sensors are of the “projected capacitance” (pCap) variety. These sensors generally lie in a planar orientation and work on the basis of mutual capacitance between a set of transmit (Tx) electrodes that run along or parallel to a first axis (for example the y-axis) and a set of receive (Rx) electrodes that run along or parallel to a second axis (for example the x-axis), and thus are arranged orthogonally to one another. While there are many different electrode patterns in use, the most common for pCap sensors is the interlocked diamond pattern shown in FIG. 4. The Tx and Rx lines are either on different layers, or they are on the same layer, but bridges are formed where the lines would otherwise intersect, so that no contact occurs between a Tx line and any Rx line.
There is a capacitive coupling from each Tx line to each Rx line, and the amount of this mutual capacitance is reduced if a finger is placed near the intersection of the two. The finger in effect “steals” some of the electric field lines that would otherwise have reached the Rx line, as represented in FIG. 5a, from Zimmerman et al., 1995. This “human shunt” is a standard mode for pCap sensing. By measuring the mutual capacitance (for which there are numerous known techniques) for each Tx-Rx pair, and interpolating the results, the centroids of the respective fingers can be located.
The same electrode pattern also can be used to measure finger locations using an older self-capacitance technique, rather than mutual capacitance. Under this approach the perpendicular lines (Rx and Tx lines) are treated equivalently. Each electrode (whether in an Rx or Tx line) has a capacitance to ground, and this capacitance is increased when a finger is brought nearby. That makes it particularly easy to detect that a finger is somewhere along any given line. X and Y coordinates are found separately by querying both the x-axis Tx and y-axis Rx electrode lines. The limitation of this approach is that it does not support multi-touch sensing very well. One must consider what happens when two fingers are placed on the touch surface. In general, two x-axis Tx lines of electrodes and two y-axis Rx lines of electrodes will respond. But those lines cross at four points, for example (x1,y1) (x1,y2) (x2,y1) (x2,y2), not two points. Two locations are correct and two other locations are misidentifications or “ghosts.” With such as system, there is not a good way to disambiguate the actual fingers from the ghosts.